This guide to writing an IEEE conference paper or technical report is published by EdCitation, the citation tool that builds each reference from the publisher's record. The layout is in IEEE paper format and the references in the IEEE referencing guide.
Every rule was read on 23 September 2026: the IEEE Author Center, the IEEE Editorial Style Manual for Authors (29 July 2024), the IEEE Reference Guide (version 3.28.2025), one conference's instructions and three universities' guides. EdCitation's Check your paper reads a call for papers or brief into a checklist, and Cite a source builds each IEEE reference from the record, both free with no account.
What are the parts of an IEEE conference paper?
Title, abstract, index terms, introduction, the body, conclusion, any appendices, acknowledgment and references, in that order (IEEE, 2024). The Author Center splits the body into methods, then results and discussion (IEEE, n.d.-d). The field adds related work and system design. Each part's job (our view where the IEEE is silent):
| Part | Its job | The common slip |
|---|---|---|
| Title | Specific, concise, descriptive | "New" or "novel"; a subtitle |
| Abstract | One standalone paragraph, up to 250 words | Citations, equations, undefined acronyms |
| Index terms | Keywords, in alphabetical order | Left out, though required |
| First footnote | Funding; any human or animal research statement | Funding put in the acknowledgment |
| Introduction | The field, the exact question, the contribution | Contributions left implicit |
| Related work | How each group of earlier work differs | One summary per paper, no comparison |
| Method or design | Enough detail to replicate the work | Versions, data or code missing |
| Results and discussion | Results, meaning and limits | "Very close" with no number |
| Conclusion | Findings, shortcomings, future work | The abstract repeated |
| Acknowledgment | Non-authors' help; AI-generated content | AI use undeclared |
| References | Only works that directly support the paper | Typed from memory, not built with EdCitation's Cite a source |
How do I write the title, abstract and index terms?
Make the title specific and the abstract a standalone account. The Author Center wants a title that is "specific, concise, and descriptive", without words such as new or novel (IEEE, n.d.-d).
The abstract
The conference page asks for one paragraph of up to 250 words on the research, its conclusions and what is unique, with no abbreviations, footnotes, references or equations (IEEE, n.d.-d). The manual sets 150 to 250 words and asks for a standalone abstract, because IEEE Xplore and indexing services display it on its own (IEEE, 2024). They differ on acronyms: the conference page allows none, the manual defines each at first use. Writing the term out satisfies both.
Index terms and the first footnote
Index terms are required, alphabetical, closing the abstract, with any acronym written out first (IEEE, 2024); the conference page asks for three to five. Funding goes in a first-page footnote, with, for research on people or animals, a statement naming the approving board or the exemption (IEEE, n.d.-d).
How do I write the introduction and state the contributions?
Move from the field to the question, then say what the paper adds. The Author Center's order is an overview of current research, the specific topic, then "the exact question or hypothesis", with the motivation and contribution stated (IEEE, n.d.-d). The Editorial Style Manual says nothing about the introduction's content, only its singular heading and drop capital (IEEE, 2024).
Stating the contributions plainly
Reviewers are asked whether the work is distinct from earlier publications and "a significant contribution to the field" (IEEE, n.d.-g). In our view, end the introduction with the contributions as claims, each pointing to its evidence: "This paper makes three contributions: 1) a [method] for [problem] (Section III); 2) an evaluation against [baselines] (Section IV); 3) [released code] (Section V)." A contribution the results never show should be cut.
Related work that compares
Our advice: group earlier work by approach and say how each group differs from yours, as in "[3] and [4] assume X; [5] removes X but needs Y; this paper needs neither." The IEEE asks for "only those references that directly support your work" (IEEE, n.d.-d) and calls citing irrelevant sources to inflate metrics a breach of ethics (IEEE, n.d.-a). EdCitation's Find sources searches about 300 million published works by topic or by claim.
A worked reference from the record
The Crossref record for DOI 10.1109/CVPR.2016.90 gives four authors, the proceedings title, Las Vegas, NV, USA, and pages 770–778. In the Reference Guide's form for proceedings with a DOI (IEEE, 2025):
[1] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. 2016 IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Las Vegas, NV, USA, 2016, pp. 770–778, doi: 10.1109/CVPR.2016.90.
The exceptions: Pattern, missing from the guide's abbreviation list, stays whole; a U.S. location ends in USA; the guide calls location optional yet says to include it if given, so include it; all four authors are named, as "et al." begins after six; and a paper presented but not published takes "presented at", not "in".
EdCitation's Cite a source, given the same DOI on 23 September 2026, returned:
K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” in 2016 IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2016, pp. 770–778, doi: 10.1109/CVPR.2016.90.
The authors, pages and DOI come from the record, and the conference's name is shortened with the Reference Guide's own word list, with no publisher and a comma before "doi:", as the guide's form has it. Two things are left to you. The title keeps the capitals the record gives it, so set it in sentence case. And EdCitation abbreviates only the words in the record's name, so add "Proc." and the location if your course follows the fuller form above.
How do I describe the method or system so it can be reproduced?
Give enough detail "for someone to replicate your work", the Author Center's own test (IEEE, n.d.-d). For a system, in our reading: the architecture, parameters, software and hardware versions, the data and its split, and the baselines.
The IEEE "strongly supports" sharing data and code, naming data repositories such as Zenodo and, for code, Code Ocean (IEEE, n.d.-c). The manual puts a dataset's DOI in the first footnote (IEEE, 2024). Under double-blind review, a repository link can name you.
How do I report results with figures and tables?
Show the results, interpret them and state their limits, without overstating (IEEE, n.d.-d). The University of Illinois's ECE 431 lab guide adds three habits (Wells et al., 2012): compare with theory by a number, such as a percentage error, never "very close"; include no figure or table the text does not mention; and give units with every value, moving unused data to an appendix.
Captions are in IEEE paper format. The IEEE bars inventing data, changing or omitting it, and altering an image's scientific meaning (IEEE, n.d.-a).
How do I write the conclusion and the acknowledgment?
State what the paper showed, where it falls short and what should come next: the Author Center asks for key findings, implications, benefits or shortcomings, and future research (IEEE, n.d.-d). The manual adds only the singular heading, Conclusion (IEEE, 2024).
The Acknowledgment
The Acknowledgment (the IEEE's spelling) follows any appendix, precedes the references, is never numbered and is written in the third person (IEEE, 2024). It thanks people who helped but are not authors (IEEE, n.d.-d).
What ethics rules does the IEEE set for authors?
Four rules bear on writing:
- Authorship. An author made a significant intellectual contribution, helped draft or revise the paper, and approved the final version; anyone else goes in the Acknowledgment (IEEE, n.d.-a).
- Plagiarism. Using someone else's work, including ideas, processes, results or algorithms, without acknowledging the source is "a serious breach of professional conduct" (IEEE, n.d.-e).
- Reuse. Recycled text from your own paper is marked and cited, and prior publications disclosed (IEEE, n.d.-e); see self-plagiarism.
- AI. AI-generated text, figures, images or code are disclosed in the acknowledgment, naming the system, the sections and how far it was used; editing and grammar help is generally outside the policy, though disclosure is recommended (IEEE, 2024). Course wording is in how to declare AI use in an assignment.
How do page limits and double-blind review work?
The conference decides both. The Author Center lists a full paper at 6 to 8 pages and says the conference website should give "full details on paper requirements", page limits included (IEEE, n.d.-f). "Most IEEE publications use the single-anonymous review format", where reviewers know the authors; in double-anonymous review, neither side knows the other. Many conferences offer no revision round (IEEE, n.d.-g).
Anonymising a submission: one conference's rules
The RFIC 2026 symposium's instructions are exact (IEEE Radio Frequency Integrated Circuits Symposium, 2025):
- Remove names, affiliations and publicly known product names from the text, cover page and PDF metadata, and logos from die photos and circuit-board images.
- Cite your own published work in the third person, "The design in [1] demonstrates", never "We demonstrated in [1]", and leave the reference unredacted.
- List your submitted, unpublished work as "Reference redacted as per double blind guidelines".
- Hold back acknowledgments and funding until acceptance; cite none of your own patents.
- Do not cite or link a preprint of the paper.
Breaking these may mean rejection without review; every submission gets a software plagiarism check.
How is a technical report or project report different?
A report is longer, opens with a summary for readers who will not read it all, and puts detail in appendices. The IEEE Professional Communication Society gives an executive summary "typically as long as 10% of the full report" and an abstract "typically less than a page long", and adds front matter such as a contents page and lists of figures and tables (IEEE Professional Communication Society, n.d.).
What three university guides ask
- University of Sussex (2022): a summary intelligible alone, typically 100 to 300 words; an introduction that does not copy the lab handout's; a bibliography of works not cited, apart from the references; appendices for code and raw data. Its citations are numbered, but it names no style.
- San José State University (2026), electrical engineering: a master's project report styled like IEEE Transactions on Computers papers; the department's cover page, or the report is rejected; an abstract of at most 150 words; derivations, algorithms or schematics in enough detail for others to continue the work. Its sample reference writes "Vol. 49" and a title-case title, unlike the current Reference Guide; follow the guide.
- University of Illinois, ECE 431 (Wells et al., 2012): IEEE referencing; cite the lab manual rather than repeat its procedure; data sheets as an appendix; third person.
The department's brief outranks all of this.
How do I get from the call for papers to the camera-ready paper?
- List the page limit, template, review type and deadlines; EdCitation's Check your paper reads the call into a checklist.
- Write the contributions, then the method.
- Find and group related work with EdCitation's Find sources.
- Build figures and tables from the final data.
- Write the results, the conclusion, then the introduction and abstract.
- Build each reference with Cite a source, in the order the paper first cites them.
- For double-anonymous review, anonymise text, figures, links and metadata.
- Draft any AI disclosure.
- Check the list with EdCitation's Verify references: every [n] matched to an entry.
- Submit, then revise if the conference allows it.
- For camera-ready, restore names and acknowledgment, delete template text, check every author's name and run PDF eXpress if asked. Nothing changes once the paper is in IEEE Xplore (IEEE, n.d.-b).
Which EdCitation tools help with an IEEE paper or report?
Check your paper is the best place to start, because the IEEE leaves the requirements to each conference (IEEE, n.d.-f), and they differ, as RFIC's rules show. Free with no account, it reads a call for papers or brief into a checklist of page or word limits, required sections, style and number of sources, each rule shown with the sentence it came from. Theoretics QA, part of Max at $24 a month, judges the paper against those instructions with quotations from it, and never rewrites anything.
Cite a source is the best tool for the list, because it looks each reference up and never writes one: the authors, pages and DOI in the worked example came straight from the publisher's record, so none of them can be invented, and the conference name comes out in the guide's abbreviations. Free with no account, it builds IEEE references from a DOI, PubMed ID, ISBN or title and screens each for retraction; since 23 September 2026 it sets "et al." in roman and an article number after the date, as the Reference Guide does. A free account keeps the list, numbered in the order you add sources, and downloads it as BibTeX for LaTeX and Overleaf, or RIS. Verify references, free, marks each entry verified, doubtful ("check this") or not found, flags retractions, never shows "could not check" as "not found", and with the paper uploaded matches every [n] to its entry.
EdCitation never writes any part of a paper and does not check the method, results or code. Mechanics QA, part of Pro at $8 a month, checks formatting against the APA 7 student paper, so it is not an IEEE format check. No tool promises acceptance or a grade.
Quick questions
How long is an IEEE conference paper?
The conference sets the limit. The IEEE Author Center lists a full paper at 6 to 8 pages and a short paper at under 4.
Does an IEEE paper structure include a related work section?
Not by rule. The IEEE puts an overview of current research in the introduction; a separate related work section is the field's convention.
Must I declare AI use in an IEEE paper?
Yes, for AI-generated text, figures, images or code, in the acknowledgment, naming the system and the sections. Editing and grammar help is generally outside the IEEE policy.
How do I cite my own work in a double-blind IEEE submission?
Cite your own published work in the third person, as "The design in [1] demonstrates", and leave the reference unredacted. That is the IEEE RFIC Symposium's rule; your conference's call may differ.
Can EdCitation check my paper against the call for papers?
EdCitation's free Check your paper reads the call into a checklist, each rule with its sentence. Theoretics QA, part of Max at $24 a month, judges the paper against the call with quotations and never rewrites.
References
- He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 770–778). IEEE. https://doi.org/10.1109/CVPR.2016.90
- IEEE. (n.d.-a). Ethical requirements. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/author-ethics/ethical-requirements/
- IEEE. (n.d.-b). Finalize your paper. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/get-published/finalize-your-paper/
- IEEE. (n.d.-c). Research reproducibility. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/write-your-paper/research-reproducibility/
- IEEE. (n.d.-d). Structure your paper. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/write-your-paper/structure-your-paper/
- IEEE. (n.d.-e). Submission policies. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/author-ethics/guidelines-and-policies/submission-policies/
- IEEE. (n.d.-f). Types of IEEE conference papers. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/become-an-ieee-conference-author/types-of-ieee-conference-papers/
- IEEE. (n.d.-g). Understand peer review. IEEE Author Center Conferences. https://conferences.ieeeauthorcenter.ieee.org/understand-peer-review/
- IEEE. (2024, July 29). IEEE editorial style manual for authors. https://journals.ieeeauthorcenter.ieee.org/wp-content/uploads/sites/7/IEEE-Editorial-Style-Manual-for-Authors.pdf
- IEEE. (2025). IEEE reference style guide for authors (Version 3.28.2025). https://journals.ieeeauthorcenter.ieee.org/wp-content/uploads/sites/7/IEEE_Reference_Guide.pdf
- IEEE Professional Communication Society. (n.d.). Write effective reports. IEEE ProComm. https://procomm.ieee.org/communication-resources-for-engineers/written-reports/write-effective-reports/
- IEEE Radio Frequency Integrated Circuits Symposium. (2025). Double-blind instructions: Updated for RFIC 2026 [PDF]. https://ims-ieee.org/sites/ims/files/content_images/RFIC/2026/rfic2026_double_blind_review_policy.pdf
- San José State University. (2026, April 20). Project and thesis guidelines. Department of Electrical Engineering. https://www.sjsu.edu/ee/graduate-program/project-and-thesis-guidelines.php
- University of Sussex. (2022, November). Guide to technical report writing. School of Engineering and Informatics. https://www.sussex.ac.uk/ei/internal/forstudents/engineeringdesign/studyguides/techreportwriting
- Wells, J., Amrhein, M., & Magill, M. (2012). A basic guide to writing a successful laboratory report [PDF]. ECE 431, University of Illinois Urbana-Champaign. https://courses.grainger.illinois.edu/ece431/sp2021/lab/ECE431_lab_report_guide.pdf